Plenary Lecture
プレナリーレクチャー |
|---|
| 7月25日（木）11:10～12:10 第1会場（朱鷺メッセ 4F 国際会議室）
1PL01
Unraveling the Mysteries of the Basal Ganglia
Ann M Graybiel（Graybiel Ann M）
The McGovern Institute for Brain Research, Massachusetts Institute of Technology (MIT)
Basal ganglia disorders can entail cognitive and affective symptoms as well as motor problems. How are these two different realms of behavior joined together? Efforts to understand these two attributes are important, as they have far-reaching influences on our behaviors. I will discuss evidence that the compartmental structure of the striatum, dividing it into striosomes and matrix, has a role in underpinning this dichotomy. Our questions include: Do striosomes, as our evidence so far suggests, have a disproportionate role in learning from cues leading to actions under challenging conditions? Can activity in one or the other of these compartments overlap to varying degrees depending on behavioral demands? How does this division relate to the D1-D2 organization of the striatum? We are using multiple tasks to explore these questions with 2-photon microscopy, fiber photometry, electrophysiology and behavior in engineered mice. Our lineage analyses suggest that the division into striosomes and matrix occurs very early, already in the ganglionic eminence. These findings suggest that there is great evolutionary conservation of striosome-matrix organization. Finally, I will point toward relating our work in non-human species to findings in the human brain in health and disease. I gratefully acknowledge the NIH, CHDI Foundation and the Saks Kavanaugh Foundation for support. | | 7月26日（金）13:00～14:00 第1会場（朱鷺メッセ 4F 国際会議室）
2PL01
Mechanisms of Synaptic Dysfunction and Loss in Alzheimer's Disease
Stephen M. Strittmatter（Strittmatter Stephen M.）
Yale University
Synapse dysfunction and loss is central to symptoms in Alzheimer's disease (AD). Oligomers of amyloid beta (Aβo) peptide trigger synaptic damage, which is accelerated by the innate immune system and linked to Tau pathology. Our work has focused on the molecular basis for Aβo interaction with post-synaptic sites to initiate this complicated pathology. Initial unbiased genome wide expression cloning studies identified cellular Prion Protein (PrPC) as a high affinity Aβo binding site. Subsequent genetic perturbation and antibody blocking demonstrated a requirement for PrPC in AD-related phenotypes in mice. Importantly, the interaction of Aβo with PrPC separates a hydrogel phase at the cell surface which engages signal transduction via the mGluR5 co-receptor in mouse models and in human AD brain. We screened for compounds blocking PrPC interaction with Aβo, identifying a range of negatively-charged polymers with low nanomolar PrPC affinity. These PrPC antagonists prevent Aβo/PrPC-hydrogel formation, block Aβo-induced synapse loss and abrogate PrPSc production. An orally available PrPC antagonist reaches the brain to rescue AD transgenic mice from AD-related synapse loss and memory deficits. Signal transduction from the Aβo-PrPC-mGluR5 complex activates the intracellular tyrosine kinases, Fyn and Pyk2, at post-synaptic sites to disrupt NMDA receptor regulation and cause Tau pathology. Pyk2 (PTK2B) genetic variation is known to contribute to late onset AD risk. We found that Pyk2 is required for synapse loss and for learning deficits in a transgenic mouse model of AD. Isolation of Pyk2-interacting proteins identified Graf1c, a RhoA GTPase-activating protein inhibited by Pyk2. Aβo-induced reductions in dendritic spine motility and chronic spine loss require both Pyk2 kinase and RhoA activation. These studies reveal a synaptic basis for Aβo-triggered synapse damage in AD, with multiple sites for potential pharmacological interruption. | | 7月26日（金）14:00～15:00 第1会場（朱鷺メッセ 4F 国際会議室）
2PL02
Neuronal circuit mechanisms for learning and memory
Andreas Luthi（Luthi Andreas）
Friedrich Miescher Institute
Classical fear conditioning is one of the most powerful models for studying the neuronal substrates of associative learning and for investigating how plasticity in defined neuronal circuits causes behavioral changes. In my talk, I will focus on the organization and function of the neuronal circuitry of fear and discuss how functionally, anatomically and genetically defined types of amygdala neurons contribute to the acquisition and expression of conditioned fear behavior. In addition to its role in conditioned fear behavior, the amygdala has long been implicated in the regulation of persistent states, such as anxiety and drive. We are using deep brain calcium imaging of genetically identified neurons while mice engage in a range of self- paced behaviors in combination with computational approaches to analyse the relation between slow dynamics at the neuronal and at the behavioral level. By tracking large neuronal populations across days and paradigms, we describe a hierarchy of activity patterns that emerges in the network dynamics during learning and corresponds to behavior on multiple timescales. Our results describe the response of the network dynamics to perturbations along different dimensions and the interplay between state-like representation and the processing of specific events and actions. Our present findings suggest a general principle of network dynamics that could underlie the involvement of the amygdala in such different functions as sensory associative learning, action selection and emotional processing. | | 7月27日（土）10:50～11:50 第1会場（朱鷺メッセ 4F 国際会議室）
3PL01
Active inference, decisions and curiosity
Karl Friston（Friston Karl）
University College London
In the cognitive neurosciences and machine learning, we have formal ways of understanding and characterising perception and decision-making; however, the approaches appear very different: current formulations of perceptual synthesis call on theories like predictive coding and Bayesian brain hypothesis. Conversely, formulations of decision-making and choice behaviour often appeal to reinforcement learning and the Bellman optimality principle. On the one hand, the brain seems to be in the game of optimising beliefs about how its sensations are caused; while, on the other hand, our choices and decisions appear to be governed by value functions and reward. Are these formulations irreconcilable, or is there some underlying imperative that renders perceptual inference and decision-making two sides of the same coin.
This talk offers a formal account of insight and learning in terms of active (Bayesian) inference. It deals with the dual problem of inferring states of the world and learning its statistical structure. In contrast to current trends in machine learning (e.g., deep learning), we focus on how agents learn from a small number of ambiguous outcomes to form insight. I will simulations of abstract rule-learning and approximate Bayesian inference to show that minimising (expected) free energy leads to active sampling of novel contingencies. This epistemic, curiosity-directed behaviour closes `explanatory gaps' in knowledge about the causal structure of the world; thereby reducing ignorance, in addition to resolving uncertainty about states of the known world. We then move from inference to model selection or structure learning to show how abductive processes emerge when agents test plausible hypotheses about symmetries in their generative models of the world. The ensuing Bayesian model reduction evokes mechanisms associated with sleep and has all the hallmarks of aha moments. | |
|
|
